Thioredoxin/glutaredoxin systems are ubiquitous in living systems from bacteria to humans, and participate in a wide variety of thiol- disulfide reactions. There are now a number of examples of metabolic control by thioredoxins, and human thioredoxin and its fragments have been strongly implicated in the inflammatory response of eosinophils and maturation of B-cells. Thioredoxin has been found at elevated levels in the serum of AIDS patients. This proposal is concerned with several important new aspects of the thioredoxin/glutaredoxin system. The first specific aim is concerned with the Escherichia coli thioredoxin system that was the main focus in the previous grant period. The source of the apparent difference in pH-dependent behavior between E. coli and human thioredoxin will be investigated by NMR studies of mutants that mimic the local sequence in human thioredoxin. In the second specific aim, a complex between a single-cysteine mutant of thioredoxin and a peptide derived from the active site of a major physiological partner, ribonucleotide reductase, (RR) will be characterized, as a mixed-disulfide model for the interaction between the two proteins. This should give important structural information on the thioredoxin binding site on RR, which is not seen in the RR crystal structure. The third specific aim represents a major focus of the work for the next grant period, the determination of the solution structure of a novel glutaredoxin, Grx2, from E. coli. This protein is large for NMR studies (Mr 24,300 Da), but preliminary spectra are outstanding. It may represent a novel fold, since it has no sequence similarity, especially $ in the C-terminal region, to any known protein. Although the function of this molecule is not yet fully understood, a knockout of the Grx2 gene is uniformly lethal to the coli bacteria. Structural information on Grx2 is a necessary prerequisite for the design of agents aimed at controlling, for example, "killer" strains of E. coli. The final specific aim is concerned with biochemical characterization and NMR structural studies of an 80-residue fragment of human thioredoxin that acts as an eosinophil cytotoxicity-enhancing factor. The full-length human thioredoxin is inactive in this system, which strongly implies that the fragment has a unique conformation, possibly dimeric, that differs fundamentally from that of the intact protein. Together these studies are focused on the long-term goal of an understanding of the mechanism of action of this important class of enzymes.